Fence around substation with no live parts

[mshields]

I've got a 5MVA substation consisting of ALL pad mounted components and full height Eaton 5kV vacuum switchgear in a walk in enclosure. As none of the components have live parts, we'd have been within code requirements not to have a fence around it. We did nevertheless call for a galvanized steel chain link fence, specifically calling for galvanized as opposed to vinyl coated such that it could be properly grounded.

What was installed was a vinyl coated. The post are grounded and tied into the ground grid but of course the chain link itself is isolated from that grounding by virtue of it's coating. I thought at first glance that the exception under 250.190 might provide an out for having to ground this. But then I see that a requirement is that the fence not be grounded. This fence is grounded to the same ground grid that grounds all of the equipment.

My opinion is that it is therefore a code violation. So a) do you agree and b) would ungrounding the fence make it code compliant? Fence posts would still be grounded via their foundation but grounding would be independent of the grounding network for the substation.

Thanks,

Mike

 

[mgookin]

Is it poco or private?

 

[don_resqcapt19]

Why do you think that fence needs to be grounded? 250.194 (new for 2014 code) does not require your fence to be grounded.

 

[Haji]

Fence posts would still be grounded via their foundation but grounding would be independent of the grounding network for the substation.

Single ground, not two grounds as above......

 

[mivey]

I've got a 5MVA substation consisting of ALL pad mounted components and full height Eaton 5kV vacuum switchgear in a walk in enclosure. As none of the components have live parts, we'd have been within code requirements not to have a fence around it.

The no live parts is only one part. Exposed metal equipment is another. If the equipment and switchgear are within an enclosure to keep someone from touching the metal then the surrounding fence does not have to be grounded and bonded.

We did nevertheless call for a galvanized steel chain link fence, specifically calling for galvanized as opposed to vinyl coated such that it could be properly grounded.

What was installed was a vinyl coated. The post are grounded and tied into the ground grid but of course the chain link itself is isolated from that grounding by virtue of it's coating.

I would doubt the insulating rating and durability of the coating. With the posts tied to the grounding grid, you should also bond the fence.

 

[mivey]

Why do you think that fence needs to be grounded? 250.194 (new for 2014 code) does not require your fence to be grounded.

Other than being enclosed what makes you think it doesn't? I saw a prior post where you mentioned no live parts as being an "out". Touch, step, and transferred voltage protection is not limited to the person touching a normally live part.

 

[don_resqcapt19]

Other than being enclosed what makes you think it doesn't? I saw a prior post where you mentioned no live parts as being an "out". Touch, step, and transferred voltage protection is not limited to the person touching a normally live part.

The code section I cited says the fence does not require grounding.

The NEC does not really take a close look at the step and touch issues, like the NESC does for substations.

If we have to address the step and touch issues around this equipment there is a lot more to look at than just the fence bonding.

 

[mshields]

Replies

Replies

I actually agree that the fence doesn't require grounding given that the exception of 250.190. But that exception indicates that I'd actually have to be isolated from ground. At present I've got the posts interconnected to the ground grid. I suppose I could just disconnect those ground connections. But the ground grid will still pass close to the posts. I think the most conservative approach at this point is to make sure the whole fence is grounded - not just the posts.

Thanks for all the input.

Mike

 

[Haji]

mshield: There is still something missing in your scheme of protection: suppose a person standing on the ground mat of substation touches the fence during a ground fault. Since fence and ground are connected and so at same potential, no problem. But if he stands outside the fence while touching it, he is likely to be electrocuted. So it an industrial practice to extend ground mat beyond fence in your case.

 

[don_resqcapt19]

mshield: There is still something missing in your scheme of protection: suppose a person standing on the ground mat of substation touches the fence during a ground fault. Since fence and ground are connected and so at same potential, no problem. But if he stands outside the fence while touching it, he is likely to be electrocuted. So it an industrial practice to extend ground mat beyond fence in your case.

It would be very unusual for the type of installation being talked about in this thread to have a ground mat.

 

[mivey]

The code section I cited says the fence does not require grounding.

Not what it states. The equipment is exposed and even though they are non–current-carrying metal parts, unless they are in a walk-in enclosure where they can't be touched, they have to be grounded.

The NEC does not really take a close look at the step and touch issues, like the NESC does for substations.

But the NEC reference for this section is exactly that.

If we have to address the step and touch issues around this equipment there is a lot more to look at than just the fence bonding.

That is what "shall be grounded and bonded to limit step, touch and transfer voltages" means.

I'm not following how you are reading it differently, especially given what the NEC states is being addressed here.

 

[mivey]

I actually agree that the fence doesn't require grounding given that the exception of 250.190.

Since people can stand on the ground and touch the fence (the "such metal parts"), that exception does not apply.

I think the most conservative approach at this point is to make sure the whole fence is grounded - not just the posts.

Very wise.:thumbsup:

 

[mivey]

It would be very unusual for the type of installation being talked about in this thread to have a ground mat.

Given that the OP said the installation he was talking about had a ground grid, I think it is very likely that the installation being talked about in this thread has a ground grid:

...and tied into the ground grid...to the same ground grid...

 

[Haji]

It would be very unusual for the type of installation being talked about in this thread to have a ground mat.

It has. But having it does not always spell that the fencing be grounded.
Because for

Single ground, not two grounds as above......

the answer is 'unless it be a 'remote' ground i.e the fence ground is not remote from ground mat i.e ground grid and within 5 feet.

 

[don_resqcapt19]

I am reading 250.194 using this definition of 'exposed"

Exposed (as applied to live parts). Capable of being inadvertently touched or approached nearer than a safe distance by a person.

The other definition of "exposed" does not apply, in my opinion, to the rule in 250.194. We do not have any live parts or equipment that can be touched in this application.

Exposed (as applied to wiring methods). On or attached to the surface or behind panels designed to allow access.

 

[mivey]

I am reading 250.194 using this definition of 'exposed"

The other definition of "exposed" does not apply, in my opinion, to the rule in 250.194. We do not have any live parts or equipment that can be touched in this application.

Thanks Don. I see where you are coming from but I don't agree that the substation fence can be ungrounded because the conductive metal is not isolated from being touched. Perhaps if it was like a class II product insulation, but it isn't. Please consider the following:

(A) Equipment Grounding. All non–current-carrying metal parts of fixed, portable, and mobile equipment and associated fences, housings, enclosures, and supporting structures shall be grounded.

Exception: Where isolated from ground and located such that any person in contact with ground cannot contact such metal parts when the equipment is energized, the metal parts shall not be required to be grounded.

So we have to ground the associated fence.

In addition to a simple ground in 250.190, we can discuss grounding AND bonding for additional reasons in 250.194. Then we consider exposed conductors and equipment.

You may take this as only applying to live parts, but when you consider the NEC references NESC design then you know that "step, touch and transfer voltages" has nothing to do with live parts but with voltage gradients and ground potential rise that may even be the result of a fault outside the substation.

We are talking about step voltages as well as touching conductive parts. The IEC has some wording that expresses the exposed-but-not-live concept:

Exposed-conductive-part (ECP): Conductive part forming part of electrical equipment which can be touched (even if out of reach) which is not live which may become live when basic insulation fails.

Also, FWIW:

Substation fences.
Conductive fences around substations shall be grounded. When a substation fence is expanded or a section is removed, fence sections shall be isolated, grounded, or bonded as necessary to protect employees from hazardous differences in electric potential.

Note to paragraph (d): IEEE Std 80-2000, IEEE Guide for Safety in AC Substation Grounding, contains guidelines for protection against hazardous differences in electric potential.

which references the exact same document as the NEC.

 

[don_resqcapt19]

Thanks Don. I see where you are coming from but I don't agree that the substation fence can be ungrounded because the conductive metal is not isolated from being touched. Perhaps if it was like a class II product insulation, but it isn't. Please consider the following:



So we have to ground the associated fence.

In addition to a simple ground in 250.190, we can discuss grounding AND bonding for additional reasons in 250.194. Then we consider exposed conductors and equipment.

You may take this as only applying to live parts, but when you consider the NEC references NESC design then you know that "step, touch and transfer voltages" has nothing to do with live parts but with voltage gradients and ground potential rise that may even be the result of a fault outside the substation.

We are talking about step voltages as well as touching conductive parts. The IEC has some wording that expresses the exposed-but-not-live concept:



Also, FWIW:


which references the exact same document as the NEC.

I understand what you are saying, but I would be willing to be that this substation does not have a grounding grid like you would find in a typical utility substation, and without that, and without the grounding grid extending beyond the fence, bonding the fence would make it more dangerous to the public outside of the fence in the event of a ground fault that resulted in ground potential rise.

 

[mivey]

I understand what you are saying, but I would be willing to be that this substation does not have a grounding grid like you would find in a typical utility substation, and without that, and without the grounding grid extending beyond the fence, bonding the fence would make it more dangerous to the public outside of the fence in the event of a ground fault that resulted in ground potential rise.

You may be right there. Perhaps mshields will elaborate.

 

[Haji]

Generally OSHA forms their rules based on NEC and other codes as below

OSHA 1926.966(d)

Substation fences.
Conductive fences around substations shall be grounded. When a substation fence is expanded or a section is removed, fence sections shall be isolated, grounded, or bonded as necessary to protect employees from hazardous differences in electric potential.

But they revise their rules on ten years period whereas NEC on three years. Hence the reason for the exception in the code not found in OSHA

NEC 250.190 Grounding of Equipment

(A) Equipment Grounding. All non–current-carrying metal parts of fixed, portable, and mobile equipment and associated fences, housings, enclosures, and supporting structures shall be grounded.

Exception: Where isolated from ground and located such that any person in contact with ground cannot contact such metal parts when the equipment is energized, the metal parts shall not be required to be grounded.

 

[don_resqcapt19]

Now we need to define "substation". I would bet that what Mike has called a substation is nothing more than a pad mount transformer and metal enclosed switchgear.

It is rare to see a pad mount transformer with any type of fencing. The switchgear would normally have fencing if the public has access, but outside metal clad switchgear on industrial sites often do not have fencing.